A Pilot Study on the Diel Vertical Migration Pattern of Mesopelagic Fishes in the Southern and Central South China Sea

Abstract

The diel vertical migration of mesopelagic fishes in the southern (S-May station and S-Nov station) and central (C-Jun station and C-Dec station) South China Sea was investigated through a series of continuous field surveys conducted in May, June, November, and December 2017. These surveys employed a combination of mid-water trawl and acoustics techniques. The diel migration process, vertical distribution, acoustic migration proportion, and migration pattern of mesopelagic fishes were analyzed. The results revealed that mesopelagic fishes initiated an upward migration before sunset, with the process concluding within 30–120 min after dark. Subsequently, they commenced a downward migration before dawn, which terminated within 10–50 min after sunrise. The mesopelagic deep-sea layers of mesopelagic fishes at the S-May, S-Nov, C-Jun, and C-Dec stations ranged from 360 to 700 m, 350 to 680 m, 350 to 520 m, and 300 to 700 m, respectively. The acoustic migration proportions of mesopelagic fishes at the corresponding stations were found to be approximately 44.5%, 25.7%, 29.8%, and 58.0%, respectively. There were seasonal and regional differences in the vertical migration and distribution patterns of mesopelagic fishes in the South China Sea. A total of 228 species were identified, including 203 fish species, 23 cephalopod species, and 2 shark species. Among these, a subset of 43 fish and cephalopod species exhibited extensive diel vertical migrating behavior. Specifically, this subset comprised 23 lanternfish species, 8 cephalopod species, and 12 other fish species. Lanternfishes were the predominant diel vertical migratory species, while cephalopods also played a significant role in diel vertical migration. The diel migration behavior was found to be slight for Diaphus lucidus, Melamphaes microps, Argyropelecus affinis, and six other fish species. Non-migratory behavior was observed in Sternoptyx obscura, Argyropelecus sladeni, Sternoptyx diaphana, and 13 other fish species. The diel migration habits of 178 additional species of fish, cephalopods, and sharks could not be definitively determined.

1. Introduction

Mesopelagic fishes predominantly inhabit the pelagic zone ranging from 200 to 1000 m and are distributed across global waters [1,2,3,4]. In a narrow sense, mesopelagic fishes refer to small fish species, such as the families Myctophidae, Stomiidae, and Gonostomatidae. However, in a broader context, mesopelagic fishes encompass not only these small fish but also cephalopods and other vertebrates, collectively known as mesopelagic nektons [5,6,7]. In the 1980s, the Food and Agriculture Organization of the United Nations (FAO) estimated the global mesopelagic fishes resource to be approximately 1 billion tons; however, recent studies indicated a potential deviation of an order of magnitude [8,9,10,11,12]. Mesopelagic fishes primarily consume zooplankton and serve as a crucial prey source for numerous higher trophic-level marine organisms, thereby playing a pivotal role in the material cycling and energy transfer within the deep-sea food web [13,14,15,16]. One of the most remarkable characteristics exhibited by mesopelagic fishes is the prevalence of diel vertical migration (DVM), wherein these fish ascend to the epipelagic surface layer for nocturnal feeding and descend to the deep sea during daylight hours for excretion [17,18,19]. Mesopelagic fishes play a crucial role in the transfer of organic carbon from the surface ocean to the deep-sea ecosystems through their feeding, metabolic processes, excretion, and vertical migration, thereby serving as a significant source of matter and energy for deep-sea ecosystems [15,20,21,22]. Currently, within the field of DVM, there are three primary approaches for specializing in the study of mesopelagic fish. The first approach involves inferring diel migration patterns by analyzing the nocturnal increase in the epipelagic surface trawl catch and decrease in the mid-layer catch. The second approach utilizes acoustic technology to detect and analyze the underwater sound scattering layer, enabling insights into the temporal and spatial distribution characteristics associated with the day and night migration of mesopelagic fish. Lastly, employing underwater submersibles equipped with optical equipment allows for the close-range observation of both diurnal and nocturnal movements of mesopelagic fish [18,23,24].

The South China Sea (SCS) is the largest marginal sea in the western Pacific Ocean, covering an area of approximately 356 × 104 km² with a mean depth of 1212 m and a maximum depth of 5500 m. The SCS boasts a high diversity and biomass of mesopelagic fishes, with recent studies expanding from morphological classification, geographic distribution, and basic biology to species composition, resource assessment, and diel vertical migration [6,16,25,26,27,28]. The investigation of diel vertical migrations among mesopelagic fishes represents a pioneering scientific inquiry in international marine ecology, playing a pivotal role in comprehending the intricacies surrounding ecosystems within this specific zone. Throughout both spring and winter 2017, we conducted continuous observations and samplings regarding vertical distributions and diel migrations of these species specifically found within southern and central regions of the South China Sea. By employing scientific echosounders alongside mid-water trawls aboard our research vessel, we successfully explored various aspects such as migratory processes throughout each daily cycle, overall spatial distributions vertically across depths, and proportions associated with migrations occurring at different timescales or distances traveled horizontally by these fish populations during their movements, all aimed towards accumulating comprehensive data necessary for achieving an enhanced comprehension regarding the structural functionality exhibited by entire ecosystems.

2. Materials and Methods

2.1. Survey Design

The time-series stations S-May and S-Nov were conducted in the southern SCS at geographical coordinates 9.5° N 113° E on 26–28 May and 21–23 November 2017, respectively (Figure 1). Similarly, the time-series stations C-Jun and C-Dec were carried out in the central SCS at geographical coordinates 14.5° N 114° E on preset dates of 6–8 June and 2–4 December 2017, respectively. Each station was continuously observed for a duration of 48 h. During the survey, the dawn–sunrise and dusk–darkness time periods at stations S-May, C-Jun, S-Nov, and C-Dec were 05:44–06:06 and 18:44–19:07, 05:31–05:54 and 18:51–19:16, 06:02–06:25 and 18:04–18:26, and 06:10–06:35 and 17:53–18:17. The fisheries resource and environmental science survey vessel Nanfeng (1537 t GT, 66.7 m in length, 12.4 m in width, and 5.0 m in draught) was chartered to conduct surveys. Throughout the observation period, the vessel primarily maintained a drifting state at a speed ranging from 0.2 to 1.0 knots, except during mid-water trawl operations. To prevent excessive deviation from the predetermined position, the vessel underwent periodic resetting approximately every four hours by activating its main engine and returning to resume drift observations.

2.2. Data Collection

The acoustic data were collected using a hull-mounted Simrad EK60 scientific echosounder (Simrad, Horten, Norway) operating at a frequency of 38 kHz (2000 W transmitting power, 1.024 ms pulse duration per 2 s, 27.06 dB transducer gain, 6.95° along. 3 dB beam width, 6.80° athward. 3 dB beam width, –20.6 dB equivalent beam angle). Before the survey, the echosounder was calibrated in accordance with international standards at the anchorage [29]. Water column hydrographic data, including temperature and salinity, were collected using a Seabird SBE-911 CTD profiler (Sea–Bird Electronics Inc., Bellevue, USA). The hydrological data were utilized for the computation of sound velocity and seawater absorption coefficient, the calibration of sonar systems, and the assessment and subtraction of ambient noise.

Mesopelagic fishes were sampled using a mid-water trawl equipped with a Simrad PI44 monitoring system (Simrad, Horten, Norway). The trawl had a straightened mouth perimeter of 136.1 m, a head tube length of 30.0 m, and consisted of four panels with codend mesh stretched to 10 mm squares, extending 20 m from the codend and measuring 40 mm in the forward section of the trawl [6]. When trawled, the mean vertical opening height was 8 m. We carried out 6 mid-water trawl operations at each station, and each haul was towed for 1 h at a vessel speed around 3.2−3.8 kn, with an average of 3.5 kn. The mid-water trawl operations are presented in

Table 1. Trawling sampling information of mesopelagic fishes in May, June, November, and December 2017 in the southern (denoted as S-May station and S-Nov station) and central (denoted as C-Jun station and C-Dec station) South China Sea.

Station Code	Sampling Time	Sampling Depth (m)	Mean Bottom Depth (m)	Sun Cycle	M/D/Y
S-May-1	04:00–05:00	75	>1000	night	05/27/2017
S-May-2	20:35–21:45	75	>1000	night	05/26/2017
S-May-3	06:30–07:30	450	>1000	day	05/28/2017
S-May-4	09:35–10:35	450	>1000	day	05/26/2017
S-May-5	13:02–14:02	550	>1000	day	05/27/2017
S-May-6	21:30–22:30	600	>1000	night	05/27/2017
C-Jun-1	19:50–20:50	75	>1000	night	06/07/2017
C-Jun-2	05:47–06:47	75	>1000	night–day	06/06/2017
C-Jun-3	14:20–15:20	500	>1000	day	06/07/2017
C-Jun-4	21:07–22:07	500	>1000	night	06/07/2017
C-Jun-5	08:07–09:07	600	>1000	day	06/08/2017
C-Jun-6	20:14–21:14	600	>1000	night	06/08/2017
S-Nov-1	22:40–23:40	75	>1000	night	11/21/2017
S-Nov-2	14:14–15:14	300	>1000	day	11/21/2017
S-Nov -3	15:26–16:26	400	>1000	day	11/22/2017
S-Nov-4	22:54–23:54	400	>1000	night	11/22/2017
S-Nov-5	07:32–08:32	500	>1000	day	11/23/2017
S-Nov-6	14:30–15:30	600	>1000	day	11/23/2017
C-Dec-1	21:00–22:00	75	>1000	night	12/04/2017
C-Dec-2	15:43–16:43	200	>1000	day	12/02/2017
C-Dec-3	09:30–10:00	300	>1000	day	12/04/2017
C-Dec-4	00:40–01:40	400	>1000	night	12/02/2017
C-Dec-5	17:50–18:50	500	>1000	day	12/03/2017
C-Dec-6	08:50–09:50	600	>1000	day	12/03/2017

.

Following trawl hauling, the catches were promptly retrieved from the trawl. Upon retrieval on deck, each trawl was emptied. Subsequently, these catches underwent immediate identification, counting, and measuring in the onboard biological laboratory. Comprehensive records were maintained, encompassing total mass, species classification, total number of species, standard length (mm), and wet body mass (0.1 g). The identification process aimed to achieve taxonomic precision by relying on morphological characteristics specific to mesopelagic fishes [30,31,32,33,34,35]. If the catch weighs less than 5 kg, a complete sample of the catch will be taken. Otherwise, after excluding all visibly larger and rare samples, the remaining samples will be selected in proportion to the weight of the catch [6].

2.3. Acoustic Data Analysis

The acoustic data were processed using the Echoview 12.0 software (Myriax Software Pty. Ltd., Hobart, Australia), following standard procedures for quality control. The processed acoustic data underwent thorough examination to identify and exclude any instances of significant transient noise, as well as to eliminate background noise interference [36,37,38]. The water depths analyzed ranged from 10 m below the water surface (5 m below the surface of the echosounder) to 1000 m. The integration threshold was set at −80 dB, and the seawater acoustic absorption for the ranges of 10–200 m and 200–1000 m was calculated based on average temperature and salinity, resulting in values of 0.006 dB/m and 0.009 dB/m, respectively [6,25,28]. The acoustic data were divided into 5 min of elementary distance sampling units and 10 m depth bins.

2.4. Vertical Distribution and Diel Vertical Migration

The nautical area scattering coefficient (NASC, m²/nmi²) was employed to analyze the acoustic vertical distribution and diel migration of mesopelagic fishes (Figure 2). The continuous variation of acoustic density over time was examined in 5 min intervals from 10 to 1000 m. Dawn, daytime, dusk, and night were represented by the periods of 04:00−07:00, 13:00−15:00, 17:00−19:30, and 22:00−24:00, respectively, at intervals of 10 m. The vertical distribution of NASC during each time period were analyzed along with evaluating the central gravity depth of the scattering layer and assessing the proportion of day and night migration [28,39]. The central gravity depth of the daytime or nighttime scattering layer was quantified as the weighted mean depth/m (WMD), which was determined through the following equation:

$$\mathrm{W}\mathrm{M}\mathrm{D}=(\sum \left({\mathrm{N}\mathrm{A}\mathrm{S}\mathrm{C}}_{\mathrm{A}\mathrm{i}}\times {\mathrm{D}}_{\mathrm{i}}\right)/\sum {\mathrm{N}\mathrm{A}\mathrm{S}\mathrm{C}}_{\mathrm{A}\mathrm{i}})$$

(1)

where NASC_Ai is the average NASC during the day or night at the 10 m depth interval. D_i is the mean depth from the ith depth interval to the sea surface. To mitigate the impact of vertical migration on diurnal and nocturnal WMD, data collected during dawn and dusk were excluded from the WMD assessment. Based on the vertical distribution characteristics observed in acoustic images, the estimated depth range of the WMD during daytime was 200–1000 m, while at night it extended from 10 to 200 m within the epipelagic surface scattering layer.

A substantial body of research has consistently demonstrated that mesopelagic fishes primarily inhabit the 200–1000 m depth range during daylight hours, with many species exhibiting upward vertical migration to the upper ocean layer (around 200 m) at nightfall. These findings strongly support the hypothesis that reduced scattering intensity observed in the mesopelagic zone (at depths ranging from 200 to 1000 m) during nocturnal periods can be attributed to organismal migration towards shallower surface layers (10–200 m) [17,18,23,28,39]. This hypothesis was further corroborated by acoustic observations, which demonstrated a migration proportion consistent with actual biomass migration from the mesopelagic deep-sea layer (200–1000 m) to the epipelagic surface layer (10–200 m) at night. The acoustic migration proportion (P_m) was determined by assessing diel variations in mean NASC, as evaluated below [28,39]:

$${\mathrm{P}}_{\mathrm{m}}=(\sum \left({\overline{\mathrm{N}\mathrm{A}\mathrm{S}\mathrm{C}}}_{\mathrm{d}\mathrm{a}\mathrm{y}200-1000\mathrm{m}}-{\overline{\mathrm{N}\mathrm{A}\mathrm{S}\mathrm{C}}}_{\mathrm{n}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}200-1000\mathrm{m}}\right)/{\overline{\mathrm{N}\mathrm{A}\mathrm{S}\mathrm{C}}}_{\mathrm{d}\mathrm{a}\mathrm{y}200-1000\mathrm{m}})\times 100\mathrm{\%}$$

(2)

where ${\overline{\mathrm{N}\mathrm{A}\mathrm{S}\mathrm{C}}}_{\mathrm{d}\mathrm{a}\mathrm{y}200-1000\mathrm{m}}$ and ${\overline{\mathrm{N}\mathrm{A}\mathrm{S}\mathrm{C}}}_{\mathrm{n}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}200-1000\mathrm{m}}$ are the mean NASC at 200–1000 m layers during daytime and nighttime, respectively.

2.5. Diel Vertical Migration Patterns

To ascertain the diel vertical migration patterns of mesopelagic fishes, we compared the compositional disparities between daytime and nighttime catches across various water layers [18,23,28,39]. Based on the diel migration behavior and its intensity, mesopelagic fishes were classified into extensively migrating species, slightly migrating species, and non-migratory species. Extensively migrating species are characterized by their daytime occupation of depths greater than 200 m and nighttime migration to shallower depths. Slightly migrating species refer to those that ascend at night but do not reach a depth of 200 m or less during the day. Non-migratory species maintain a consistent distribution depth throughout both day and night periods. Only species exclusively sampled at night in the epipelagic surface layers were considered as exhibiting extensive diel vertical migration habits.

3. Results

3.1. Migration Process

According to the acoustic density variation, mesopelagic fishes at the southern station in May 2017 (S-May) began migrating upward successively at 17:17 (86 min before sunset). The extensive migration started at 18:22 and concluded by 19:37. The extensive downward migration commenced at 05:25 (18 min before dawn) and ended at 06:52, marked by the stabilization of fish acoustic density in the layers of 10–200 m and 200–1000 m (Figure 2). The mesopelagic fishes began their upward migration at 18:00 (50 min before sunset) in June 2017 (C-Jun). The extensive upward migration started at 18:35 and essentially ended by 21:17. Subsequently, a significant downward migration occurred from 04:50 (50 min before dawn) until it finished at 06:15 (Figure 3).

In November and December 2017, the upward migration of mesopelagic fishes at the southern station (S-Nov) commenced at 17:47 (6 min prior to sunset), with the extensive ascent process essentially concluding by 19:17. Subsequently, the substantial descent migration initiated at 04:50 (70 min before dawn) and terminated by 06:37 (Figure 3). The ascent of mesopelagic fishes at the central station (C-Dec) commenced at 17:05 (48 min prior to sunset), with the upward migration process predominantly ending by 18:40. Subsequently, an extensive descent occurred from 04:57 onwards (73 min before dawn), terminating at 06:57 (Figure 4).

In brief, the mesopelagic fishes at stations S-May, C-Jun, S-Nov, and C-Dec exhibited a sequential upward migration before sunset, with the large-scale ascent process generally concluding within 30–120 min after dark. Subsequently, they initiated a downward migration towards the mesopelagic deep-sea layer prior to dawn. The large-scale descent was completed within 10–50 min after sunrise.

3.2. Vertical Distribution

The diel vertical distribution of mesopelagic fishes showed seasonal and spatial variations, as observed in the acoustic density profiles (Figure 5 and Figure 6). At the S-May station, mesopelagic fishes were primarily found at depths between 360 and 700 m (WMD: 530 m) during the day. However, after their evening migration, these fishes predominantly occupied surface layers ranging in depth from 10 to 130 m (WMD: 60 m). At night, a mesopelagic deep-sea scattering layer was observed at depths between 340 and 700 m with two sub-layers: one at depths of 340–500 m (WMD: 440 m) and another at depths of 550–700 m (WMD: 640 m). The mesopelagic fishes at C-Jun station were mainly distributed at depths of 350–520 m (WMD: 465 m) and 550–780 m (WMD: 670 m) during the day. The nighttime surface scattering layer ranged in depth from 10 to 160 m (WMD: 75 m), while the mesopelagic scattering layer consisted of two sub-layers, one around depths of 380–520 m (WMD: 450 m) and the other around depths of 560–750 m (WMD: 660 m).

The mesopelagic fishes were mainly found at depths of 350–680 m during the day at the S-Nov station, with a WMD of 490 m, which was lower compared to May. At night, the epipelagic surface scattering layer extended from 10 to 150 m, with a WMD of 80 m, larger than that observed in May. Additionally, the mesopelagic deep-sea scattering layer encompassed depths of 360–480 m and 530–740 m, with WMDs of 450 m and 660 m, both slightly greater than those recorded in May. The mesopelagic fishes at C-Dec station were mainly observed at depths of 300–700 m (WMD: 660) during the day. At night, the epipelagic surface scattering layer was found at shallower depths ranging from 10 to 140 m, with a WMD of 50 m, which was smaller compared to June. The mesopelagic deep-sea scattering layer consistently occurred at depths of 340–490 m, with a WMD of 440 m; this differed from its distribution pattern in June.

The mesopelagic fishes migrated downwards from the epipelagic surface layer to the mesopelagic deep-sea layer and upwards from the mesopelagic layer to the epipelagic surface layer during two distinct periods: 04:00–07:30 and 17:00–19:30, respectively. As a result, the vertical distribution of acoustic density during these time intervals shows characteristics of both diurnal and nocturnal patterns.

3.3. Acoustic Migration Proportion

The diurnal average NASCs at the S-May station were 5501 ± 54 m²/nmi² and 2413 ± 200 m²/nmi² in the epipelagic surface and mesopelagic deep-sea layers, respectively, while those were 2305 ± 263 m²/nmi² and 1339 ± 290 m²/nmi² at night, respectively, and the P_m was 44.5%. At the C-Jun station, the diurnal average NASCs were 261 ± 68 m²/nmi² and 2864 ± 71 m²/nmi² in the epipelagic surface and mesopelagic deep-sea layers, respectively, while those were 1807 ± 71 m²/nmi² and 2010 ± 152 m²/nmi² at night, respectively, and the P_m was 29.8%. The NASC of the mesopelagic deep-sea layer at the S-May station was lower than that at the C-Jun station, whereas the migration proportion exhibited a higher value compared to the latter.

At the S-Nov station, the average NASCs were 436 ± 116 m²/nmi² and 2602 ± 178 m²/nmi² in the epipelagic surface and mesopelagic deep-sea layers during the day, respectively, while those were 2186 ± 138 m²/nmi² and 1934 ± 56 m²/nmi² at night, respectively, and the P_m was 25.7%. At the C-Dec station, the average NASCs were 402 ± 86 m²/nmi² and 3270 ± 450 m²/nmi² in the epipelagic surface and mesopelagic deep-sea layers during the day, respectively, while those were 2092 ± 721 m²/nmi² and 1372 ± 454 m²/nmi² at night, respectively, and the P_m was 58.0%. The NASC in the mesopelagic deep-sea layer during the day and the P_m were lower at the S-Nov station than at the C-Dec station. Seasonal variations were observed in the NASC and migration proportion of mesopelagic fishes at the two stations.

3.4. Diel Vertical Migration Species

A total of 228 species were identified, including 203 fish species, 23 cephalopod species, and 2 shark species. The extensive diel vertical migration habit was observed in a total of 43 species of fish and cephalopods, with lanternfishes accounting for the majority (53.5%) at 23 species, followed by cephalopods (18.6%) with 8 species, and other fish (27.9%) with 12 species (

Table 2. Extensive migratory mesopelagic fishes species in the southern and central South China Sea.

Order	Family	Species	Distributed Depth		Occurrence
			Min.	Max.
1	Myctophidae	Diaphus fulgerns	75 m (N)	550 m (D)	10/24
2	Myctophidae	Lampadena mollis	75 m (N)	600 m (D)	14/24
3	Myctophidae	Diaphus phillipsi	75 m (N)	600 m (D/N)	16/24
4	Myctophidae	Diaphus fragilis	75 m (N)	600 m (D)	19/24
5	Myctophidae	Diaphus garmani	75 m (N)	600 m (D/N)	22/24
6	Myctophidae	Diaphus perspicillatus	75 m (N)	600 m (D/N)	4/24
7	Myctophidae	Diaphus chrysorhynchus	75 m (N)	550 m (D)/600 m (N)	12/24
8	Myctophidae	Diaphus malayanus	75 m (N)	600 m (D)	16/24
9	Myctophidae	Diaphus brachycephalus	75 m (N)	600 m (D/N)	15/24
10	Myctophidae	Diaphus watasei	75 m (N)	550 m (D)	6/24
11	Myctophidae	Diaphus jenseni	75 m (N)	550 m (D)	5/24
12	Myctophidae	Lampanyctus hubbsi	75 m (N)	600 m (D)	8/24
13	Myctophidae	Lampanyctus nobilis	75 m (N)	550 m (D)	4/24
14	Myctophidae	Lampanyctus omostigma	75 m (N)	600 m (D/N)	10/24
15	Myctophidae	Lampanyctus festivus	75 m (N)	600 m (D)	7/24
16	Myctophidae	Lampanyctus punctatissimus	75 m (N)	600 m (D)	8/24
17	Myctophidae	Bolinichthys longipes	75 m (N)	600 m (D/N)	21/24
18	Myctophidae	Myctophum asperum	75 m (N)	600 m (D)	7/24
19	Myctophidae	Lampadena luminosa	75 m (N)	600 m (D)	6/24
20	Myctophidae	Symbolophorus evermanni	75 m (N)	600 m (D/N)	9/24
21	Myctophidae	Hygophum proximum	75 m (N)	600 m (D/N)	19/24
22	Myctophidae	Centrobranchus andreae	75 m (N)	600 m (D)	7/24
23	Myctophidae	Ceratoscopelus warmingii	75 m (N)	600 m (D)	12/24
24	Notosudidae	Scopelosaurus hoedti	75 m (N)	550 m (D)/600 m (N)	6/24
25	Paralepididae	Stemonosudis rothschildi	75 m (N)	600 m (D)	3/24
26	Gonostomatidae	Vinciguerria nimbara	75 m (N)	600 m (D)	10/24
27	Gonostomatidae	Diplophos taenia	75 m (N)	550 m (D)/600 m (N)	8/24
28	Nomeidae	Cubiceps squamiceps	75 m (N)	600 m (D)	4/24
29	Nomeidae	Cubiceps pauciradiatus	75 m (N)	600 m (N)	3/24
30	Stomiidae	Astronesthes splendidus	75 m (N)	―	2/24
31	Chauliodontidae	Chauliodus sloani	75 m (N)	600 m (D)	16/24
32	Grammicolepidae	Xenolepidichthys dalgleishi	75 m (N)	600 m (D/N)	8/24
33	Idiacanthidae	Idiacanthus fasciola	75 m (N)	600 m (D)	6/24
34	Nomeidae	Psenes arafurensis	75 m (N)	450 m (D)	4/24
35	Bramidae	Brama myersi	75 m (N)	600 m (D)	4/24
36	Enoploteuthidae	Abralia similis	75 m (N)	600 m (D)	3/24
37	Enoploteuthidae	Abratia multihamata	75 m (N)	500 m (D)	2/24
38	Enoploteuthidae	Abraliopsis atlantica	75 m (DW)	―	1/24
39	Enoploteuthidae	Abratia andamanica	75 m (N)	600 m (D)	13/24
40	Pyroteuthidae	Pterygioteuthis giardi	75 m (N)	600 m (D/N)	2/24
41	Mastigoteuthidae	Mastigoteuthis cordiformis	75 m (N)	600 m (N)	6/24
42	Ommastrephidae	Ornithoteuthis volatilis	75 m (N)	600 m (D/N)	7/24
43	Cranchiidae	Leachia pacifica	75 m (N)	500 m (D)/600 m (N)	4/24

Note: In the table, 75 (N), 75 (DW), and 550 (D) stand for the distribution depths of species at night, at dawn, and during the day (75 m, 75 m, and 550 m), respectively. In the occurrence column, 10/24 denotes 10 occurrences of the species in 24 nets.

). Six fish species, including Diaphus lucidus, Melamphaes microps, and Argyropelecus affinis, were identified as exhibiting slight diel vertical migration habits, constituting 2.6% of the total species. Among them, one was a lanternfish while the remaining five belonged to other fish taxa (

Table 3. Weak migratory mesopelagic fishes species in the southern and central South China Sea.

Order	Family	Species	Distributed Depth		Occurrence
			Min.	Max.
1	Myctophidae	Diaphus lucidus	200 m (TW)	600 m (D)	9/24
2	Melamphaidae	Melamphaes microps	400 m (N)	600 m (D)	3/24
3	Sternoptychidae	Argyropelecus affinis	200 m (N)	600 m (D)	13/24
4	Gonostomatidae	Gonostoma gracile	400 m (N)	600 m (D)	5/24
5	Gonostomatidae	Margrethia obtusirostra	450 m (DW)	600 m (D)	5/24
6	Scopelarchidae	Scopelarchus analis	400 m (N)	600 m (D)	4/24

Note: In the table, 200 m (TW), 400 m (N), 450 m (DW), and 600 m (D) stand for the distribution depths of species at twilight, at night, at dawn, and during the day (200 m, 400 m, 450 m, and 600 m), respectively. In the occurrence column, 9/24 denotes 9 occurrences of the species in 24 nets.

). Thirteen fish species were non-migratory species, constituting 5.7% of all species (

Table 4. Non-migratory mesopelagic fishes species in the southern and central South China Sea.

Order	Family	Species	Distributed Depth		Occurrence
			Night	Day
1	Sternoptychidae	Sternoptyx obscura	500–600 m	450–600 m	6/24
2	Sternoptychidae	Argyropelecus sladeni	400–600 m	450–600 m	8/24
3	Sternoptychidae	Sternoptyx diaphana	400 m	400–600 m	10/24
4	Melamphaidae	Melamphaes polylepis	400–600 m	400–600 m	13/24
5	Melamphaidae	Melamphaes lugubris	500–600 m	400–600 m	7/24
6	Gonostomatidae	Vinciguerria attenuata	500 m	450–600 m	7/24
7	Gonostomatidae	Cyclothone pseudopallida	400–600 m	450–600 m	11/24
8	Evermannellidae	Evermannella indica	600 m	450–550 m	5/24
9	Phosichthyidae	Woodsia nonsuchae	400–500 m	400–600 m	4/24
10	Stomiidae	Heterophotus ophistoma	500 m	600 m	4/24
11	Serrivomeridae	Serrivomer beanii	400–600 m	400–600 m	7/24
12	Malacosteidae	Malacosteus niger	500–600 m	500–600 m	7/24
13	Scopelarchidae	Scopelarchus guentheri	400 m	400 m	5/24

Note: In the occurrence column, 6/24 denotes 6 occurrences of the species in 24 nets.

). Among the 228 species, only 62 species (27.2% of the total) could be unequivocally identified, while the remaining 166 fish, cephalopod, and shark species lacked clear determination. Notably, non-migratory species exhibit an absence of lanternfishes and cephalopods.

4. Discussion

4.1. Diel Vertical Distribution

Since the 1960s, investigations on diel vertical distribution patterns of mesopelagic fishes have been conducted in various marine regions, confirming the prevalence of vertical migration behaviors among many mesopelagic species [1,5,10,18,20]. During daytime, trawling operations are ineffective in capturing mesopelagic fishes within the shallow water layer at a depth of 200 m; however, they can successfully target these species within the deep-sea scattering layer ranging from 200 to 1000 m. Notably, acoustic scattering intensity is considerably weak in the shallow water layer at a depth of 200 m during daylight hours, while it becomes significantly enhanced during nighttime when mesopelagic fishes occupy this region. Conversely, acoustic scattering signals within the 200–1000 m water column become notably weaker at night. With the advancement of observational technology and the comprehensive investigation content, research has revealed that nektons such as fish and cephalopods play a significant role in the DSL, specifically within the context of mesopelagic fishes. The periodic diel migration exhibited by migratory species in the mid-ocean and epipelagic surface oceans during twilight periods is a prevalent and distinctive characteristic observed among open-ocean mesopelagic fishes; however, variations exist in terms of vertical distribution patterns and migration behaviors across different marine regions [20,23,39]. Liu et al. [27] conducted continuous observations on the diel vertical migration of the DSL in the waters of 12°24′ N 117°42′ E in the SCS, revealing the presence of two prominent scattering layers primarily composed of mesopelagic fishes. Between 16:00 and 19:00, certain fish species within the depth range of 350−700 m gradually ascended towards the surface layer (0−100 m), while during the period from 04:00 to 07:00, these migratory individuals descended back to depths ranging between 350 and 700 m.

In the northern slope of the SCS, vertical migration behavior on a large scale typically occurred during twilight periods, both before and after sunset and sunrise [28]. Mesopelagic fishes were predominantly distributed near the surface down to approximately 150 m at night, exhibiting a WMD of 79 m in October 2014 and a WMD of 66 m in June 2015, while during the daytime, they primarily occupied depths ranging from 300 to 800 m (being further divided into two main sub-layers: 300−550 m and 600−800 m) with a WMD of 490 m in October 2014 and a WMD of 450 m in June 2015 [28]. In our study, the WMDs of the nocturnal surface scattering layer in the southern and central South China Sea differed from those of the corresponding scattering layers in the northern continental slope of the SCS; however, the WMD in the mesopelagic deep-sea layer at the southern station in December 2017 remained consistent with that observed in October 2014 on the northern slope of the SCS. The P_m values of mesopelagic fishes in the northern continental slope of the South China Sea were 55% and 44% during October 2014 and June 2015, respectively [28]. The P_m values of mesopelagic fishes at the southern stations in May and November 2017 and central stations in June and December 2017 were 45% and 30%, and 26% and 58%, respectively. The average P_m of mesopelagic fishes (43%) observed in the South China Sea was lower than that recorded for the western Pacific region (62%) [39].

4.2. Diel Vertical Migration Pattern

The DVM behavior of mesopelagic fish significantly influences the interspecies distribution patterns within their habitat depth. During daylight hours, most migratory species predominantly inhabit the deep sea; however, during dusk (approximately one hour before and after sunset), they actively utilize their swimming abilities to overcome the barrier effects imposed by halocline and thermocline layers spanning several hundred meters vertically. Consequently, they engage in a prolonged period of vertical migration lasting several hours, with speeds ranging from 10 to 200 m/h, in order to access shallower water layers primarily within a depth range of 0–200 m. Throughout the night, migratory species continue their activities near the surface and return to deeper waters at dawn (around one hour before sunrise) [21,40,41,42,43].

The DVM of mesopelagic fishes is a crucial process for material transport and energy flow in oceanic ecosystems; however, variations exist in the vertical distribution and diel migration patterns of mesopelagic fishes across different marine regions and biological communities. Depending on the occurrence and intensity of DVM, mesopelagic fishes can be categorized into distinct modes including extensive migration, slight migration, and non-migration. Nevertheless, the migratory behavior of mesopelagic fishes may exhibit variability due to growth stages, hydrological factors, and ocean topography [17,18,20,28,44,45]. Olivar et al. [5] investigated the diel migration patterns of mesopelagic fishes in the Mediterranean and observed that the family Gonostomatidae, predominantly comprising slightly migrating or non-migratory species, primarily occupies the 400−600 m water layer. The extensively migrating lanternfishes, such as those belonging to the genus Ceratoscopelus, inhabit the 200−600 m water layer during daylight hours and ascend to shallower depths at night. However, a few larger bodied lanternfishes from genera like Lampanyctus still inhabit deeper water layers during nocturnal periods.

In the northern continental slope of the SCS, a total of 92 fish species (excluding cephalopods) exhibited diel migration patterns, including both extensive and slight migrations. Among these, lanternfish comprised 47 species (19 belonging to the Diaphus genus), accounting for 51.1% of the observed migratory species [28]. In our study, a total of 40 fish species were identified as exhibiting diel migration behavior, with lanternfish comprising the majority (23 species, including 12 belonging to the genus Diaphus), accounting for 57.5% of the observed migratory species. This proportion slightly exceeds that found in the northern continental slope.

In May 2017, the mesopelagic deep-sea scattering layer at S-May during the nighttime ranged from 340 to 700 m and exhibited two distinct sub-layers: 340–500 m (with a WMD of 440 m) and 550–700 m (with a WMD of 640 m), respectively. The NASC observed in these sub-layers can be attributed to two plausible factors: firstly, the presence of non-migratory species that inhabit this water layer throughout the day; and secondly, the upward migration of less migratory species from deeper water layers to occupy this specific water layer [7]. For migratory species, certain studies categorize them as either incomplete migrants or complete migrants based on the congruence of their habitat water layer before and after migration [46,47]. Incomplete migratory species, such as Gonostomatidae, Sternoptychidae, and Melamphaidae species (also known as slight migratory species), exhibit overlapping resident water layers both before and after migration. These species typically inhabit deep waters. Complete migratory species are characterized by their robust migratory ability, wherein their habitats undergo complete separation pre- and post migration. This phenomenon is exemplified in the case of most lanternfishes [4,40,46,47]. Complete migratory fish also exhibit variations in the extent of their vertical migration paths. Taking lanternfish as a prime example, their vertical migration encompasses the transition from deep-water layers to the sea surface, whereas Chauliodontidae fish refrain from migrating towards shallower depths [47,48].

The initiation and termination times of DVM behavior exhibit variations across diverse water depths, seasons, and cohorts [44,45,48]. Furthermore, certain mesopelagic fishes exhibit plasticity in their migratory behavior to adapt to physiological demands and fluctuations in the marine ecosystem; for instance, Vinciguerria nimbaria undergoes surface dwelling for extended periods during gonadal development [49]. However, the determination of slightly migrating and non-migratory species is prone to error or controversy due to their limited capture in the epipelagic surface layers during nighttime. Additionally, incomplete sampling across a wide range of spatial and depth dimensions leads to an inadequate representation of species composition within the catch. Simultaneously, it also encounters challenges related to fish evasion from trawls [10]. Furthermore, it is crucial to enhance the sampling frequency and refine the automated stratified trawling technique in future investigations on diurnal vertical migration patterns of mesopelagic fishes, as determining the distribution depth of different classes during various time periods remains a challenging task.

There are also limitations in our study regarding the vertical distribution and migration of mesopelagic fishes, particularly with regard to a potential under-representation of results. Given the vast area (3.5 million km²) and large latitude span (approximately 2000 km longitudinally and 1000 km east–west) of the SCS, only two stations were selected for observing and analyzing these aspects. Therefore, it is important to note that our findings may not fully capture the characteristics of mesopelagic fishes’ vertical distribution and migration in the SCS. Another significant limitation pertains to assessing migration proportions using acoustic methods. While this approach is commonly employed internationally, it should be acknowledged that changes in acoustic scattering intensity can also be influenced by zooplankton contributions, which presents a challenging limitation that currently cannot be completely eliminated. Future research could potentially enhance evaluation accuracy by considering techniques such as acoustic frequency difference.

4.3. Influencing Factors of DVM

DVM represents a strategic survival mechanism employed by mesopelagic fish populations, characterized by highly active behavior [22,50]. DVM is primarily motivated by foraging and predator avoidance, although it can also be influenced by environmental factors.

4.3.1. Biological Factors

Mesopelagic fishes primarily feed on zooplankton [13,16]. It is hypothesized that these migratory species surface feed at night and deep-sea forage during the day, with larger scale foraging observed when they ascend from the depths at night. For species exhibiting DVM, relying solely on deep-sea feeding does not provide sufficient energy for their daily growth. The increased energy obtained through nocturnal surface feeding activities after migration surpasses that gained through DVM consumption, supporting the inference that baiting plays a crucial role in inducing DVM in mesopelagic fishes.

The predation pressure on mesopelagic fish primarily comes from visual predators [40,50]. During the day, the ocean surface is illuminated, attracting large predators and causing zooplankton to migrate deeper. As a result, mesopelagic fish tend to inhabit the deep ocean during this time due to low temperature and oxygen levels that limit high-nutrient predators with specific requirements. This ensures food security for mesopelagic fish and reduces their vulnerability to natural enemies. At night, when predator activity decreases, abundant food resources are available as zooplankton aggregate near the sea surface; thus, mesopelagic fish actively feed in these areas. By minimizing exposure to predators while meeting their nutritional needs through DVM, populations ensure survival.

4.3.2. Environmental Factors

Light is the primary factor influencing mesopelagic fishes’ DVM, determining their timing, depth, and speed of migration [4,7,10]. Mesopelagic fish have adaptations for light attenuation and the avoidance of intense illumination [4]. During daylight hours, when strong light and high levels of ultraviolet radiation prevail at the sea surface, mesopelagic fish mainly inhabit the DSL and adjacent water layers [5,18,20]. At night, however, they exhibit higher density near the surface, with a shallower distribution in deeper waters. Light intensity also influences both upper and lower trophic level activities within the mesopelagic ecosystem. Visual hunting is crucial for most mesopelagic predators, making their activity patterns and hunting success rate directly dependent on ambient light conditions.

In addition to light, temperature and salinity significantly influence the DVM of mesopelagic fish [15,19,49]. The presence of water masses with diverse physical and chemical properties in the ocean contributes to the high spatial heterogeneity of the marine environment, impacting the species composition and spatial distribution patterns of mesopelagic fish. Fish with broader temperature tolerance ranges have wider migration ranges. Wang et al.‘s study revealed that some migratory groups experience maximum temperature differences up to 20 ℃ between different water layers [28]. Chauliodus displays distinct vertical migration patterns between tropical and temperate populations; while tropical populations rarely migrate below 500 m, temperate populations can be found at depths around 200 m [48]. Salinity affects mesopelagic fish migration from two perspectives: (1) dynamic changes in salinity across different water bodies impact salinity tolerance among mesopelagic fishes; and (2) salinity influences water turbidity by affecting nutrient levels. Mesopelagic fishes species with broader salinity tolerance ranges are capable of migrating along longer paths, resulting in larger DVM ranges in low-salinity areas [51].

5. Conclusions

The mesopelagic fishes exhibited a sequential upward migration before sunset, with the large-scale ascent process generally after dark. Subsequently, they initiated a downward migration towards the mesopelagic deep-sea layer prior to dawn. Our study implied that light could strongly influence the migration timing of mesopelagic fishes in the SCS. The vertical migration and distribution patterns of mesopelagic fishes in the SCS exhibited seasonal and regional variations. The acoustic migration proportions of mesopelagic fishes in the SCS were also significantly lower compared to those observed in the western Pacific region. Given the extensive latitudinal range of the SCS, it is imperative to investigate the DVM patterns of mesopelagic fish across its entire expanse in order to comprehend their DVM regularity. In our study, lanternfish were identified as the primary mesopelagic fish group exhibiting DVM, which aligns with findings from previous studies. Out of the 228 species surveyed, only 43 exhibited clear evidence of extensive vertical migration behavior. Further comprehensive sampling and analysis are imperative to elucidate migration patterns across a broader range of species, as this constitutes fundamental data for investigating the ecological significance of mesopelagic fishes in the carbon cycle within the SCS.